Numerous network data unit handling schemes have been developed to improve the performance of data unit capture and transmission. For example, software improvements relating to kernel performance for moving data units from the kernel space to the user space have been developed. Alternatively, hardware improvements, directed to, for example, interrupt mitigation and prioritization have been developed. Despite these improvements, latency stemming from context switching, locking and other synchronizations may still exist, reducing the efficiency of transferring data units through a network.